| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2000;70:1313-1318
© 2000 The Society of Thoracic Surgeons
a Lankenau Institute For Medical Research, Lankenau Hospital, Wynnewood, Pennsylvania, USA
b Division of Thoracic and Cardiovascular Surgery, Department of Surgery, Lankenau Hospital, Wynnewood, Pennsylvania, USA
Address reprint requests to Dr Marano, Cytokines and Inflammation, Lankenau Institute for Medical Research, 100 Lancaster Ave, Wynnewood, PA 19096
e-mail: carusos{at}msn.com
Abstract
Background. Tumor necrosis factor-
(TNF) has been implicated in the development of postoperative morbidity after cardiopulmonary bypass for myocardial revascularization. Despite their postulated roles as modulators of TNF bioavailability, soluble TNF receptors have not been characterized in patients undergoing this procedure and is the focus of this study.
Methods. Soluble tumor necrosis factor receptor I (sTNFRI) and TNF were measured by immunoassay in plasma samples collected from 36 patients at events before, during, and after cardiopulmonary bypass.
Results. Plasma concentrations of sTNFRI averaged 1.39 ng/mL at the start of the operation. Preoperative sTNFRI concentrations were found to significantly correlate with a preoperative morbidity assessment score, age, duration of bypass, duration of supplemental oxygen, and length of hospital stay. Plasma sTNFRI increased in all of the patients during the procedure. Plasma concentrations of sTNFRI and TNF did not correlate at any time.
Conclusions. Preoperative measurement of sTNFRI could potentially serve as a reliable indicator for prophylactic treatment with an anti-TNF therapy. Such a therapeutic approach might help attenuate inflammatory processes thought to underlie postoperative morbidity associated with cardiopulmonary bypass.
Cardiopulmonary bypass (CPB) for myocardial revascularization results in an inflammatory response that is characterized by the activation of white blood cells, adhesion molecules, and the complement cascade [1, 2]. Systemic concentrations of circulating proinflammatory cytokines, including tumor necrosis factor- (TNF), have been variably reported to be unaffected or increased in patients undergoing this procedure [3, 4]. The discrepancy in measured TNF concentrations may in part arise from its complicated chemistry in solution as well as by the presence of its own soluble receptors (sTNFRs) that are generated by the cleavage of the membrane-associated TNF receptors (I and II) by a metalloproteinase [5] and can function as binding proteins. Both sTNFRI and sTNFRII have been detected in the blood after elevations in TNF and serve either as agonist or antagonist in modulating the bioavailability of circulating TNF [6, 7]. In contrast to TNF, which is a relatively short-lived and chemically complex molecule with numerous interfering factors to obscure its reliable measure, sTNFRs have a longer half-life and a much less complicated protein chemistry, making their immunodetection a potentially more reliable indicator of inflammatory activation [8]. Therefore, in this study we set out to determine the plasma concentrations of sTNFRI in patients undergoing CPB for the purpose of myocardial revascularization. We also investigated potential associations between measured concentrations of sTNFRI and a patient’s progress through and recovery from the operation.
Material and methods
Patients
Thirty-six nondiabetic patients undergoing elective CPB for myocardial revascularization at the Lankenau Hospital from August 1996 through October 1998 were enrolled in the study and are characterized in Table 1. Before the operation, patients undergoing CPB were evaluated for their potential morbid and mortal outcomes using the scoring system based on weighted preoperative risk factors developed by Higgins and associates [9] at the Cleveland Clinic. No patient enrolled in this study was intentionally placed on drugs that could alter his or her immune response. All subjects gave informed consent, and the study was conducted after approval of the hospital’s institutional review board for human subject protocols.
|
 |
 |
 |
 |
 |
Blood collection
Whole blood samples were obtained from the radial arterial line and the arterial filter line at fixed events associated with the CPB procedure: (1) after placement of lines before induction of anesthesia; (2) at cannulation; (3) 15 minutes after application of the cross-clamp; (4) 1 minute after bypass; (5) 5 minutes after administration of protamine; (6) just before leaving the operating suite; and (7) 1 day postoperatively in the surgical intensive care unit. All samples were collected into sterile Vacutainer tubes containing 15% potassium EDTA (Becton Dickinson, Franklin Lakes, NJ) and kept on ice pending separation of plasma. Plasma was obtained by centrifugation (1,000 rpm for 10 minutes at room temperature) in an IEC-HN SII table-top centrifuge (International Equipment Co, Needham Heights, MA). The plasma was aliquoted into sterile polypropylene tubes and stored at -70°C until assayed for TNF and sTNFRI. All measurements were made on freshly thawed aliquots.
Tumor necrosis factor- and soluble tumor necrosis factor receptor I determinations
Plasma immunoreactive TNF concentrations were determined using a commercially available sandwich enzyme-linked immunoabsorbent assay kit (Dupont NEN, Boston, MA) per the manufacturer’s instructions. All samples were run undiluted. Addition of the soluble receptors (sTNFRI and sTNFRII) to the assay were found not to interfere with the measurement of TNF (data not shown). Interassay variation was 8%, and intrassay variation averaged 9%.
Concentrations of sTNFRI in the plasma were measured by an enzyme-linked immunoabsorbent assay supplied by R & D Systems (Minneapolis, MN) per the manufacturer’s instructions. Plasma samples were diluted 10-fold in diluent provided by the manufacturer. Interassay and intrassay variation both averaged 3%.
Statistical analysis
Preoperative TNF concentrations clearly identified two groups of patients. The separation was confirmed by a full normal plot, which identified 200 ng/mL TNF as the cut-off between the two groups of patients. Comparisons of sTNFRI and TNF concentrations for events 1 to 6 were performed using the two-way analysis of variance for repeated measures. Logarithms of all TNF and sTNFRI concentrations were used in the analysis of variance, and antilogarithm averages were developed for Tables 2 and 3,respectively. The monotonic increasing trend for plasma sTNFRI as a function of events was confirmed (p = 0.006) in this analysis. Average sTNFRI concentrations for event 7 were compared using the one-way analysis of variance for the logarithms. Logarithmic sTNFRI concentrations were used in all calculations for the Pearson correlation coefficients.
|
|
Results
Tumor necrosis factor- in patients undergoing cardiopulmonary bypass
Measurement of plasma concentrations of immunoreactive TNF before the induction of anesthesia (event 1) revealed two distinct groups among the 36 patients under study. Tumor necrosis factor- concentrations either fell above or below 200 pg/mL (26 versus 10; p < 0.001). Of these patients followed through the procedure, 32 manifested no meaningful change in plasma TNF concentrations relative to measurements made at event 1. These include a group (n = 22) of patients with TNF concentrations less than 200 pg/mL (72 ± 48 pg/mL) and a group (n = 10) with average TNF concentrations more than 200 pg/mL (604 ± 486 pg/mL). The remaining 4 patients experienced a 5- to 10-fold increase in plasma TNF detectable at event 4, 1 minute after bypass (Table 2; low-high group B). This elevation in plasma immunoreactive TNF concentration reached its maximum immediately after bypass and remained elevated (relative to concentrations measured at event 1) up to 24 hours after the operation, when the last sample was collected. Average TNF plasma concentrations for each of the three groups (low–low, low–high and high–high) are presented for all seven events in Table 2.
Soluble tumor necrosis factor receptor I in patients undergoing cardiopulmonary bypass
Analysis of the plasma samples collected before the induction of anesthesia (event 1) yielded an average sTNFRI concentration of 1.39 ng/mL for the 36 patients enrolled in this study. The preoperative concentrations of sTNFRI were independent of the distribution of TNF above or below 200pg/mL at the start of the operation. In addition, all patients, regardless of their intraoperative TNF profile, showed a monotonic increase in the concentration of sTNFRI during the operation, as shown in Table 3. However, by 24 hours, the plasma concentrations of sTNFRI were decreasing in each patient group. After accounting for the three subgroups of patients identified by their distinct profile of TNF as a function of time, there was no statistically significant correlation between TNF and sTNFRI concentrations. However, patients who underwent an increase in TNF during the operation also tended toward higher concentrations of sTNFRI in plasma samples collected preoperatively before the induction of anesthesia (event 1; Table 3).
Immunoassay of sTNFRII was also conducted in a subset of these plasma samples and showed a similar profile to sTNFRI although at slightly higher absolute concentrations (data not shown). The appearance of both of these sTNFRs in the plasma are supportive of the induction of proinflammatory signals during the bypass procedure.
Association of soluble tumor necrosis factor, receptor I and tumor necrosis factor, with preoperative risk assessment and outcome in patients undergoing cardiopulmonary bypass
When preoperative plasma concentrations of sTNFRI were analyzed for their association with preoperative morbidity assessment, age, duration on bypass and cross-clamp, time spent on supplemental oxygen, and length of hospital stay, statistically significant correlations were noted (Table 4). The correlation of sTNFRI with hospital stay was also significant (r = 0.641; p < 0.001) when the summed total (all events) concentration of the soluble receptor was used for analysis.
|
|
Comment
Although coronary artery bypass grafting together with CPB is generally a well-tolerated surgical procedure, it is associated with the development of varying degrees of postoperative morbidity, such as pulmonary edema, that complicate and prolong the recovery period and increase the cost of care [10]. In approximately 2% of the CPB population, a multiorgan dysfunction or postperfusion syndrome develops that manifests as noncardiac pulmonary edema or acute respiratory distress syndrome, occasionally presenting with impaired gastrointestinal permeability and renal failure that results in significant postoperative mortality among those thus affected [11, 12].
Activation of a systemic inflammatory response is an acknowledged consequence of CPB. The components of the response include activation of peripheral and resident white blood cells and the complement cascade and systemic elevations of both proinflammatory and antiinflammatory cytokines [1–4, 13]. Much effort has been placed on the effect of CPB on the systemic concentrations of the proinflammatory cytokine, TNF. Many of the postoperative complications of CPB resemble those seen with septic shock, for which lipopolysaccharide-induced elevations of TNF and its soluble receptors have been implicated [14]. Our own work with epithelial cell cultures has shown that the multiorgan complications associated with septic shock, as well as those arising after CPB, might in part be explained by the increased transepithelial permeability that results from TNF exposure [15, 16]. These observations, together with the causative effect that TNF plays in the development of pulmonary edema [17, 18], focused our efforts on the associations between TNF and its binding protein, sTNFR, to postoperative outcome after coronary artery bypass grafting accompanied by CPB. This is important as the presence of sTNFR in the blood can modify the biologic activity of TNF.
Both agonistic and antagonistic actions have been attributed to the sTNFRI binding protein in the regulation of TNF bioactivity [6, 7]. Soluble receptors to TNF are generated by the cleavage of the membrane-bound TNF receptors (I and II) by a metalloproteinase, an event that follows after TNF binds to its membrane-associated receptors [5]. After elevations in TNF, both sTNFRI and sTNFRII can be immunologically detected [14]. Initially viewed as antagonists of TNF [5], the soluble receptors have now been documented to prolong the half-life of TNF by reducing its susceptibility to proteolysis and reducing its rate of clearance, thus conferring an agonist function to their mechanism of action [7, 19]. Binding to sTNFRI may prolong the half-life of TNF but it may also have an immediate antagonistic action by inhibiting the interaction of TNF with cell surface receptors [7, 19]. A rising ratio of sTNFRI/TNF may therefore indicate that in the short term, circulating TNF is being effectively neutralized even while its biologic half-life may be extended. Elevated sTNFR concentrations in general have been shown to be an indication of a systemic inflammatory response, with the elevation stemming from a prior induction and secretion of TNF itself [20]. In contrast to TNF, which is a relatively short-lived and chemically complex molecule with numerous interfering factors to obscure its reliable measure [8], sTNFR has a longer half-life and has a much less complicated protein chemistry, making its immunodetection a more reliable quantitative indicator of an inflammatory response.
Although the patients in this study came from an elective surgical population and overall presented with a low incidence of postoperative complications, most of these patients did develop some degree of increased lung water, which typically resolved itself. None of these patients progressed to acute respiratory distress syndrome, and none died as a result of the operation. The majority of these patients did not manifest an operation-induced increase in plasma TNF concentrations, although a few experienced a 5- to 10-fold rise in systemic concentrations. With regard to sTNFRI, however, they all underwent a monotonic increase, strongly suggesting an inflammatory response to the procedure as it progressed. This apparent discrepancy between measured concentrations of TNF, its soluble receptor, and the response to bypass is not without precedent as it has been reported for numerous other nonsurgical medical conditions [14, 19].
The ultimate objective of this study was to determine whether an association existed among TNF, its soluble receptor, and a patient’s postoperative recovery from the CPB procedure. If such an association was identified, it might suggest the potential application of anti-TNF therapies in the management of the cardiac surgical patient. To this end, we found that sTNFRI concentrations significantly correlated with the time a patient spent on supplemental oxygen after the procedure, as well as his or her length of postoperative hospital stay. The former association suggests that activation of an inflammatory response (reflected as elevated sTNFRI) resulted in impaired pulmonary gas exchange that required the patient to supplement his or her oxygen intake. This condition may be the result of poor gas exchange owing to the varying degrees of increased lung water or atelectasis that can arise as a consequence of the procedure. In addition, such preoperative characteristics as age and Cleveland Clinic Score, as well as intraoperative conditions such as the duration of bypass and cross-clamp, were also found to positively correlate with preoperative sTNFRI plasma concentrations.
Whereas preoperative concentrations of sTNFRI were found to associate with a number of preoperative, intraoperative, and postoperative characteristics of the patients undergoing CPB and to increase as a patient progressed through the procedure, this was not the case for TNF. Preoperative concentrations of TNF were not found to correlate with any patient characteristic. Furthermore, only a small number of patients in this study experienced any significant change in TNF concentrations with the operation. The few patients who did undergo a 5- to 10-fold increase in plasma TNF concentrations were the oldest, spent a longer time on bypass and cross-clamp, required a longer duration of supplemental oxygen, and remained in the hospital the longest. Their small number, however, precludes a meaningful demonstration of statistical significance. The data suggest that an intraoperative increase in plasma TNF concentrations is more likely to occur in older patients who will then be in the operating room longer and have a longer recovery.
In summary, these results have important implications for the medical management of patients about to undergo CPB. They indicate that it is the inflammatory state of the patient at the start of the operation that brings the greatest weight to bear on a his or her progress through the surgical procedure as well as postoperative recovery. Furthermore, these data, unlike a patient’s age, identify a potential target for preoperative intervention in the form of anti-TNF pharmacologic therapy (eg, thalidomide and its analogs [21]) or immunotherapy (eg, anti-TNF monoclonal antibody or TNF receptor fusion proteins [22, 23]) to ameliorate or prevent certain postoperative complications such as increased lung water, thereby hastening patient recovery and shortening hospital stay. Unlike the proposed use of these anti-TNF therapies in chronic inflammatory diseases, their use here would be one-time or at most short-term, and therefore be of even less expense and lower risk. The advantage to the patient would be the avoidance of certain postoperative complications by interfering very early in the inflammatory cascade rather than later when the number of proinflammatory mediators and their associated damage is greater.
Acknowledgments
The authors thank Karen Shearman and Sophia Ingram of Main Line Cardiothoracic Surgeons for their untiring efforts in the collection of the clinical variables characterizing each patient in this study. We also thank Mike Clancy, John Mitchell, and Alexander "Phil" Garwood, perfusionists, for directing the collection of the samples during the operation, as well as the anesthesia team for collecting the samples in the operating room. We are indebted to S. Michael Free, PhD, for his assistance with the statistical analysis of the data. We would also like to thank the nurses of the CTICU for their patience in the collection of the day 1 postoperative sample. We are also grateful to the editorial staff of the Institute for Medical Research for their assistance in the preparation of this manuscript, and Dr Alejandro Peralta Soler for his critical reading of the manuscript. Finally, we gratefully acknowledge the support of the Lankenau Foundation and the National Institutes of Health to James M. Mullin (CA48121) during the course of this study.
References
in patients undergoing cardiopulmonary bypass. Am J Hematol 1995;49:237-239.[Medline]
increases sodium and chloride conductance across the tight junction of CACO-2 BBE, a human intestinal epithelial cell line. J Membr Biol 1998;161:263-274.[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
